Offical Pytorch implementation of Remote Sensing papar Deep Hash Remote-Sensing Image Retrieval Assisted by Semantic Cues
An network architecture, which can classify and retrieve remote-sensing images under a unified framework, and the classification labels are further utilized as the semantic cues to assist in network training.
A standard hash code structure which integrate the classification results into the hash-retrieval process to improve accuracy.
- Python3
- Pytorch(>1.0)
- Numpy
- tqdm
- wandb
- Pytorch-Metric-Learning
Download public benchmarks
Note that a sufficiently large batch size and good parameters resulted in better overall performance than that described in the paper. Larger hash code size means better results with greater consumption.
The Evaluation Metrics is MAP@20
| Method | Backbone | 16bits | 32bits | 48bits | 64bits |
|---|---|---|---|---|---|
| Ours | Inception-BN | 98.97 | 99.34 | 99.54 | 99.60 |
The Evaluation Metrics is MAP@20
| Method | Backbone | 16bits | 32bits | 48bits | 64bits |
|---|---|---|---|---|---|
| Ours | Inception-BN | 94.75 | 98.08 | 98.93 | 99.02 |
Our code is modified and adapted on these great repositories:
If you use this method or this code in your research, please cite as:
@article{
title = {Deep Hash Remote-Sensing Image Retrieval Assisted by Semantic Cues},
journal = {Remote Sensing},
volume = {14},
pages = {2358},
year = {2022},
doi = {https://doi.org/10.3390/rs14246358},
url = {https://www.mdpi.com/2072-4292/14/24/6358#},
author = {Pingping Liu and Zetong Liu and Xue Shan and Qiuzhan Zhou},
keywords = {Remote sensing, Image retrieval, Deep hash, Metric learning},
abstract = {With the significant and rapid growth in the number of remote-sensing images, deep hash methods have become a research topic. The main work of deep hash method is to build a discriminate embedding space through the similarity relation between sample pairs and then map the feature vector into Hamming space for hashing retrieval. We demonstrate that adding a binary classification label as a kind of semantic cue could further improve the retrieval performance. In this work, we propose a new method, which we called deep hashing, based on classification label (DHCL). First, we propose a network architecture, which can classify and retrieve remote-sensing images under a unified framework, and the classification labels are further utilized as the semantic cues to assist in network training. Second, we propose a hash code structure, which can integrate the classification results into the hash-retrieval process to improve accuracy. Finally, we validate the performance of the proposed method on several remote-sensing image datasets and show the superiority of our method.}
}